• No results found

Association of hydrophobic organic compounds to organic material in the soil system

N/A
N/A
Protected

Academic year: 2021

Share "Association of hydrophobic organic compounds to organic material in the soil system"

Copied!
53
0
0

Loading.... (view fulltext now)

Full text

(1)

Association of hydrophobic

organic compounds to organic

material in the soil system

Silviu-Laurentiu Badea

Department of Chemistry

Licentiate Thesis Umeå 2013 Umeå University

(2)

Copyright©Silviu-Laurentiu Badea ISBN: 978-91-7459-594-9

Front Cover Photo: Thomas Liljedahl

Electronic version available on http://umu.diva-portal.org/ Printed by: VMC, KBC, Umeå University

(3)
(4)
(5)

Table of Contents

List of papers iii

Abbreviations v

Abstract vii

1. Introduction 1

1.1. Objectives 4

2. Materials and Methods 5

2.1. Selected Hydrophobic Organic Compounds 5

2.1.1. Hexachlorocyclohexans (HCHs) 5

2.1.2. Polychlorinated phenols (PCPh) 6

2.1.3. Polychlorinated benzenes (PCBz) 7

2.1.4. Polycyclic Aromatic Hydrocarbons (PAHs) 7

2.1.5. Polychlorinated Biphenyls (PCBs) 8

2.1.6. Polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) 10

2.2. Soil properties 11

2.3. Leaching tests 12

2.3.1. Factors affecting leachability 13

2.3.1.1. pH 14

2.3.1.2. Organic Matter (OM) 14

2.4. Characterization of soil organic matter and matrices 15

2.4.1. X-ray Photoelectron Spectroscopy (XPS) 15

2.4.2. Fourier transform infrared spectroscopy (FTIR) 16

3. Results and Discussions 19

3.1. Variation of leachability in relation to soil composition and dissolved dissolved organic carbon 19 3.2. Qualitatively characterization of soil organic matter and dissolved organic carbon 25 3.3. Variation of the distribution coefficients with hydrophobicity of the compounds 28 3.4. Correlation between leachability of ortho-PCBs and DOC 30

4. Concluding remarks and future aspects 32

4.1. Conclusions 33

4.2. Future aspects 34

Acknowledgements 36

(6)
(7)

List of Papers

This thesis is based on the following papers, which are referred to in the text by their respective Roman numerals.

I. Badea, S.-L.; Lundstedt, S.; Liljelind, P.; Tysklind, M., The influence of soil composition on the leachability of selected hydrophobic organic compounds (HOCs) from soils using a batch leaching test. Journal of Hazardous Materials 2013, 254–255, 26-35.

II. Badea, S.-L.; Mustafa, M.; Lundstedt, S.; Tysklind, M., Leachability of PCBs in soil and its pH and dissolved organic matter (DOM) dependency. Manuscript

Paper I is reprinted with permission from Elsevier.

Contribution by the author of this thesis to the papers

Paper I: The author was highly involved in the planning of the experiment, performed the whole

experimental work, and wrote the paper.

Paper II: The author was highly involved in the planning of the experiment, performed most of

(8)

Paper published by author but not included in this thesis

I. Badea, S. L.; Vogt, C.; Weber, S.; Danet, A. F.; Richnow, H. H., Stable Isotope Fractionation of gamma-Hexachlorocyclohexane (Lindane) during Reductive Dechlorination by Two Strains of Sulfate-Reducing Bacteria. Environmental Science & Technology 2009, 43, (9), 3155-3161. II. Badea, S. L.; Vogt, C.; Gehre, M.; Fischer, A.; Danet, A. F.; Richnow, H. H., Development of an enantiomer-specific stable carbon isotope analysis (ESIA) method for assessing the fate of alpha-hexachlorocyclohexane in the environment. Rapid Communications in Mass Spectrometry 2011, 25, (10), 1363-1372.

(9)

Abreviations

BaA Benzo(a)anthracene

DL-PCBs Dioxin like - polychlorinated biphenyls

FTIR Fourier transform infrared spectroscopy

GC-HRMS Gas chromatography-high resolution mass spectrometry

HCBz Polychlorinated benzenes

HCHs Hexachlorocyclohexan

HOCs Hydrophobic organic compounds

I-PCBs Indicator PCBs

m-o PCBs Mono-ortho PCBs

DOC Dissolved organic carbon

DOM Dissolved organic matter

Kd Soil-water distribution coefficient Kow Octanol-water partition coefficient

LAS Linear alkylbenzene sulfonate

NDL-PCBs Non-dioxin like polychlorinated biphenyls

OCs Organochlorine pesticides

OPLS Orthogonal projections to latent structures

PAHs Polycyclic aromatic hydrocarbons

PCBs Polychlorinated biphenyls

PCDDs/Fs Polychlorinated dibenzo-p-dioxins/furans

PCPhs Polychlorinated phenols

Phe Phenanthrene

(10)

TEF Toxic equivalent factor

TOC Total organic carbon

(11)

Abstract

Contaminated soils and sediments have been identified as significant secondary sources of organic contaminants. Leaching tests may be useful tools to estimate the mobility of contaminants via the water phase and thereby the risk for groundwater and surface water contamination. The influence of soil composition (peat and clay content) on the leachability was investigated in batch leaching experiments for chemically diverse hydrophobic organic compounds (HOCs: PCP, PAHs, HCB, HCHs, PCBs, and TCDD/Fs). The above mentioned compounds were analyzed by both GC-LRMS (gas chromatography coupled with low resolution mass spectrometry (GC-HRMS) and GC-HRMS (gas chromatography coupled with high resolution mass spectrometry). Also the the leachability of eleven selected PCBs from naturally aged soil (Västervik, Sweden) was investigated in relation to the composition and concentration of dissolved organic matter at different pH (2 to 9), using a pH static test with initial acid/base addition. The the composition and of dissolved organic matter (DOM) at different pH values was explored by FTIR spectroscopy. The results were evaluated by orthogonal projections to latent structures (OPLS).

Generally, for all model compounds studies, the Kd-values showed a variability of 2-3 orders of

magnitude depending on the matrix composition. The Kd-values of moderately hydrophobic compounds,

(e.g. HCHs, PCP and Phe), were correlated mainly with the organic matter content of soil. For more hydrophobic compounds (e.g.BaA, HCB and PCB 47), the leachability decreased as the proportions of OM and clay contents increased. The Kd-values of 1,3,6,8-TCDD and 1,3,6,8-TCDF were positively

correlated with peat content but negatively correlated with clay content, while for PCB 153 and PCB 155 the correlations were reversed. The log Kd-values of all target PCBs decreased with increased pH values

and the log Kd-values were highly correlated with the concentration of total organic carbon (TOC) in the

leachates. The FTIR analysis of DOM showed that the least chlorinated and hydrophobic PCB congeners (i.e. PCB 28) might be associated with the hydrophilic fraction (i.e. carboxylic groups) of DOM. Our study demonstrated how complex interaction between the organic matter, clay components, pH and DOC influences the leachability of HOCs in a compound-specific manner.

(12)

1. Introduction

During the last two centuries, the industrialization has led to large-scale releases of industrial chemicals worldwide. Many of these chemicals can be classified as persistent organic pollutants (POPs), and once released, they are persistent in the environment, bioaccumulate in organims and potentially harmful for humans and wildlife. However, during the last decades, increasing environmental awareness has led to reductions in the release of many of these chemicals and some restrictions and bans of the production and use of certain chemicals have been taken. In many cases the contaminants have persisted for more than a century, since many of them are resistant to biodegradation and often highly hydrophobic. Contaminated soils and sediments have thus been identified as significant secondary sources of contaminants. Tarnowskie Gory (Poland), Bitterfeld (Germany) and Port of Rotterdam (Netherlands) are among the most well-known contaminated mega sites in Europe. The total number of contaminated sites in Europe is estimated to as many as 3 million sites of which more than 240.000 sites are estimated to be an object for remedial actions. At many sites, a complex contamination situation with both heavy metals and persistent organic pollutants which pose great environmental risks as they may function as secondary sources of hazardous compounds. Consequently, during the coming decades, extensive remediation activities have to be taken. Nevertheless, even using the advanced technologies of 21st century, the remediation of contaminated sites is extremely costly, in many cases costing in the order of millions of euros per site. Thus, in an attempt to decrease the costs of remediation processes, cost-efficient remediation methods have to be developed. Prior to remediation activities, a risk assessment must be performed, in which migration pathways of contaminants must be considered and estimated (see Fig. 1). The mobility of contaminants is a crucial factor in the assessment procedure and closely connected to the risks associated to a specific contamination situation. In this respect, leaching tests may be useful tools to estimate the mobility of contaminants via the water phase and thereby the risk for groundwater and surface water contamination [1]. Generally, all leaching tests aim to determine the fraction of contaminants that are loosely bound and therefore may be mobilized into the water phase. The leachability of the compounds is dependent on the inherent physico-chemical properties of the compounds, the soil type, the presence of other contaminants, as well as the age of the

(13)

contamination. How these factors will affect the specific situation cannot be estimated on the basis of theoretical knowledge alone due to the complexity of the interaction of the factors.

Fig 1. Sorption and desorption processes on the surface of soil particles are determining the mobility

pathways of HOCs in the soil compartment (in part from Jonsson et al. 2009 [1]). The present thesis investigate the leachability in relation to soil composition (Paper I) and relation to pH (Paper II)

Dust

Sediment

Separate phase

Surface water

Gas

Groundwater

Migration via water Other means of migration

Fine soil structure

Soil composition

(Paper I)

DOM and pH

(Paper II)

(14)

Thus knowledge of the total concentration in soil may not give a proper indication of the actual leachability at a specific site. Among the above mentioned factors, the soil composition and the physico-chemical properties of the compounds are the most important. Soil is composed of many different components, of which clays, oxides and organic matter are reported to be the primary constituents responsible for the sorption of organic contaminants [2]. Organic matter has a high affinity for many non-polar compounds and is considered to have a dominating influence on the sorption process. For soils and sediments with low organic matter content, clay minerals may have a more important role [3]. When it comes to the properties of the contaminants, their water solubility (Sw) as well as their partitioning behavior between water and organic phases, often

described by the octanol-water partition coefficient (Kow), are thought to be the most important

[4]. The most hydrophobic compounds (Kow > 6) are unlikely to leach in a truly dissolved phase,

due to their low water solubility and high affinity for particle surfaces. In contrast, less hydrophobic compounds (Kow < 6) are generally weakly sorbed and will leach to varying degrees

depending on their solubility.

pH is an important parameter, in relation to leaching of organic compounds from soil. This is partly due to dissociation of organic acids and to the fact that dissolved organic carbon (DOC) generally is strongly dependent on pH and it is known that different fractions of organic acids will go into solution at different pH values, with fulvic acids being released at lower pH values and humic acids at pH 7 and higher. In addition, pH greatly affect the sorption of ionisable organic compounds, e.g. chlorophenols, since the pH affects not only the surface characteristics of natural solids (e.g. surface charge and potential) but also the speciation of compounds. At neutral pH, the overall sorption has been found to be dominated by sorption of the dissociated fraction [5, 6]. Only a few papers dealt with the influence of pH on soil sorption coefficients of contaminants. Carboxylic groups in humic matter will show various degrees of dissociation when the pH changes from 4 to 7. The dissociated species have no H-bond donor property anymore and undergo completely different electrostatic interactions. While this change might have little effect on non-polar contaminants, it may significantly influence sorption of polar chemicals [7]. Nevertheless, these pH effects have not yet been studied systematically.

(15)

1.1. Objectives

The overall aim of this thesis was to investigated the sorption and desorption of a wide range of hydrophobic organic compounds (HOCs) in different soil matrixes using different types of leaching test. The soil organic matter and matrices were characterized using different types of spectroscopic techniques (FTIR, XPS , X-ray Diffraction).

The thesis is based on two papers (Paper I and II) and has the following objectives:

 To investigate the influence of the soil composition on the leachability of a wide range of HOCs, focusing on similarities and differences between the compounds.

 To identify the moieties in the organic matter responsible for interaction with the HOCs  To investigated the log Kd vs. log Kow relationship for less hydrophobic compounds and

more hydrophobic compounds

 To investigate the leachability of selected PCBs from naturally aged soil in relation to the composition and concentration of dissolved organic matter (explored by FTIR analysis) at different pH using a pH static test with initial acid/base addition.

 To study and describe the DOC-dependent leachability of selected PCBs

 To establish pH-specific Kd-values based on the concentration of PCBs at different pH

(16)

2. Materials and Methods

2.1. Selected Hydrophobic Organic Compounds

Hydrophobic organic compounds (HOCs), often present in the contaminated soils, include a wide array of compounds, e.g. organochlorine pesticides (OCs), polychlorinated phenols (PCPh) polychlorinated benzenes (PCBz), polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs), polychlorinated dibenzofurans (PCDFs), and polychlorinated dibenzo-p-dioxins (PCDDs). Among these HOCs classes the following model compounds were included in the study underlying Paper I: hexachlorocyclohexanes (HCHs), pentachlorophenol (PCP), hexachlorobenzene (HCB), phenanthrene (Phe), benzo(a)anthracene (BaA), PCB 47, PCB 153 and PCB 155, 1,3,6,8-tetrachloro dibenzo-p-dioxin (1,3,6,8-TCDD) and 1,3,6,8-tetrachloro dibenzofuran (1,3,6,8-TCDF). In Paper II, eleven PCB congeners were selected as model compounds.

2.1.1. Hexachlorocyclohexans (HCHs)

Heavy use of organochlorine insecticides has led to the dispersal of these pollutants throughout the global environment. Among them, the HCHs are of great concern. Technical-grade lindane typically contains contain 60−70% α-HCH, 5−12% β-HCH, 10−12% γ-HCH, 6−10% δ-HCH, and 3−4% ε-HCH [8]. All HCH isomers are toxic and are also considered to be carcinogenic [9] and thus of major concern for human health; those compounds can persist for years in the environment [10]. The persistence and recalcitrance of each HCH isomer is attributed to the orientation of the chlorine atoms on the molecule, which can be axial (‘a’) or equatorial (‘e’) (See Fig. 2). It is thought that axially oriented chlorine atoms are more available for enzymatic attacks than equatorial chlorine atoms. Thus, -HCH having three axially oriented and three equatorially oriented chlorine atoms (‘aaaeee’) is more easily biodegraded than -HCH (‘aeeeee’) or -HCH (‘eeeeee’) [10]. Based on the results of a number column leaching tests performed in different labs using soils of both high and low organic carbon content as well as municipal refuse, it was concluded that HCHs have generally a low mobility in soils [11-13] . Generally, it is believe that that adsorption of HCHs to soil particulates is more significant than their leaching to groundwater. However, Nordmeyer et al. [14] concluded that aquifer

(17)

sediments, which have generally a low organic carbon content, are not sufficient to absorb HCHs to the extent that groundwater contamination is prevented.

Fig. 2. Chemical structure of % α, β, and γ HCH (from Wiberg et al ). 2.1.2. Polychlorinated phenols (PCPh )

Polychlorinated phenols consist of a hydroxylated benzene ring substituted by one to five chlorine atoms (Fig. 3).

Fig. 3. Generalized structural formula and substitution positions of PCPh.

Of the 19 PCPh congeners possible, 14 are substituted in at least one ortho-position, i.e. the carbon site adjacent to the hydroxyl group. The hydroxyl group greatly increases the polarity of

(18)

the PCPh molecule relative to that of the corresponding polychlorinated benzene (PCBz) compound [16]. Therefore, the PCPh also possesses higher reactivity and water solubility comparing with corresponding PCBz, however, the polarity decreases with increasing chlorination [16]. Nevertheless, the fully-chlorinated pentachlorophenol (PCP) (the only target PCPh used in Paper I) can be found in waters dissociated into its phenolate ion (PCP-), if the pH values

are above 7 [17, 18]. The PCP has been extensively used as fungicides and for wood conservation purposes.

2.1.3. Polychlorinated benzenes (PCBz)

Polychlorinated benzenes is a group of chloroaromatic compounds, composed of a benzene ring with one to six chlorine atom substituents, resulting in a total of 12 possible congeners (Fig. 4).

Fig. 4. Generalized structural formula and substitution positions of PCBz.

The fully chlorinated species, HCB, formerly used as synthetic feedstock in industrial chemical processes and as an agricultural fungicide was selected as one of the model compounds in Paper I. HCB was one of the ten chermicals included the Stockholm convention of POPs in 2001 . Nevertheless, nowadays, the occurrence of HCB in the environment is largely due to its formation as a by-product in various industrial processes.

2.1.4. Polycyclic Aromatic Hydrocarbons (PAHs)

PAHs is a group of chemicals that is formed during the incomplete combustion of coal, oil, gas, wood, garbage, or other organic substances, such as tobacco and charbroiled meat. Theoretically, hundreds of different PAHs can be formed, but some are more commonly found in the environment than others. PAHs generally occur as complex mixtures (for example, as part of combustion products such as soot), not as single compounds. PAHs consist of fused benzene rings in linear, angular or clustered arrangements (Fig. 5), and contain by definition only carbon

(19)

and hydrogen atoms. PAHs have been thoroughly studied due to their toxicity, persistency and environmental prevalence [21]. Usually, the studies have been limited to 16 PAHs, designated as priority pollutants by the United States Environmental Protection Agency (US-EPA) [22].

Fig. 5. Structures of the 16 US-EPA PAHs (from Lundstedt et al. [23]).

2.1.5. Polychlorinated Biphenyls (PCBs)

Polychlorinated biphenyls belong to one class of pollutants with the chemical formula C12H10−nCln, where n is the number the chlorine atoms varying from 1 to 10.

(20)

Depending on the number of chlorine atoms attached to two rings, 209 congeners of PCBs are theoretically possible. However, only approximately 130 congeners have been identified in commercial PCB products [4]. These specific congeners are named according to Cl atoms position on the two rings of biphenyls, respectively [24]. IUPAC names and the corresponding substitution patterns of the target PCBs included in Paper II are listed in Tab.1.

Tab. 1. IUPAC number and IUPAC names for selected PCBs studied in Paper II.

No. IUPAC no. IUPAC name

1 28 2,4,4’-Trichlorobiphenyl 2 52 2,2’,5,5’-Tetrachlorobiphenyl 3 101 2,2’,4,5,5’-Pentachlorobiphenyl 4 118 2,3’,4,4’,5-Pentachlorobiphenyl 5 138 2,2’,3,4,4’,5’-Hexachlorobiphenyl 6 153 2,2’,4,4’,5,5’-Hexachlorobiphenyl 7 180 2,2’,3,4,4’,5,5’-Heptachlorobiphenyl 8 66 2,3’,4,4’-Tetrachlorobiphenyles 9 105 2,3,3',4,4'-Pentachlorobiphenyl 10 156 2,3,3’,4,4’,5-Hexachlorobiphenyl 11 187 2,2',3,4',5,5',6-Heptachlorobiphenyl

*Congeners 1-7 are defined as so-called indicator PCBs [25].

The toxicological properties of PCBs are related to their substitution pattern. PCBs that are lacking or only have one chlorine atom in the ortho-positions (2,2’,6 or 6’), can rotate freely around the biphenyl bond and adopt a planar conformation. Among these sub-group of PCBs, twelve congeners have a chlorine substitution pattern leading to a planar structure similar to the PCDD/Fs and therefore they can also bind to the aryl hydrocarbon receptor (AhR). Their similar toxicological pattern, compared to PCDD/Fs, has also motivated the assignment of WHO-TEF values [26] and thus twelve co-planar congeners are known as the ‘dioxin-like’ PCBs or

(21)

WHO-PCBs. Four of them are congeners that lack chlorines in the ortho-position, the non-ortho PCBs (77, 81, 126, 169), and eight have one chlorine substituent in the ortho position, the mono-ortho PCBs (105, 114, 118, 123, 156, 157, 167, 189). The ortho-substitution also affects some physico-chemical properties of PCBs that are important for their mobility, for instance, PCB hydrophobicities and log Kow values tend to decrease with increased number of ortho-chlorines

[27, 28] .

From environment point of view, many studies have been focused on seven PCB congeners , the so-called Indicator PCBs (I-PCBs, congeners nos. 1 – 7 in Tab. 1). These I-PCBs are abundant and widespread compounds in the environment and animal tissues and markers of the major components of the technical mixtures of PCBs used.

2.1.6. The polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs)

Fig. 6. General chemical structure of PCDDs (left) and PCDFs (right).

The polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs), commonly referred to as ‘dioxins’ and ‘furans’, are planar compounds that can have 1-8 chlorine substituents (Fig. 6). They constitute groups of 75 and 135 congeners, respectively, or in total 210 congeners. PCDD/Fs have never been commercially produced. Instead, they are formed as by-products in various combustion processes such as waste incineration and biomass burning, in the production of ferrous and nonferrous metals, in the bleaching of pulp with chlorine gas, and

(22)

in the production of various chlorinated chemicals [29]. Due to improvements in the cleaning technology, the industrial emissions of PCDD/Fs have been drastically reduced [29], and the relative importance of sources such as open burning of waste and biomass [30] or secondary sources such as the release from contaminated sites [31] has increased.

2.2. Soil properties

In Paper I, OECD standard soil, containing 70% sand, 20% kaolin clay and 10% sphagnum peat [32], was used as a reference soil during batch leaching experiments. In addition to the OECD soil, seven other artificial soils, with different proportions of the same constituents were prepared (see Paper I). Kaolinite (Al2Si2O5(OH)4) was identified as the main mineral in the clay, as

determined by X-ray Diffraction (Rietveld's method) [33], while the quartz (SiO2) was identified

as the main mineral (45 %) in the sand (see Tab. 2). In order to identify the moieties in the organic matter responsible for interaction with the HOCs, the peat used in the preparation of the soils was investigated by X-ray photoelectron spectroscopy (XPS) [34]. Some key properties of OCED soil used in Paper I are given in Tab.2.

Tab. 2. The soils used in Paper I and Paper II

Type of soil Kaolinite (%) Quartz (%) Water content (%)

Lost of Ignition (%) at 5500C

OECD soil 12.7 45 21.7 11.9

Västervik soil 55 10 61.0 37.5

In Paper II, contaminated soil sampled from Västervik (Sweden) contaminated site was used to perform the pH static test. By X-ray Diffraction (Rietveld's method), the main mineral phase of this soil was identified as Kaolinite (55 %), while the quartz was presented just in a small proportion (10%) (see Tab. 2). The water content of the soil was very high (63 %), while the lost of ignition (LOI) determinated at 550 0C was 37.5 %, indicating a high carbon content.

(23)

2.3. Leaching tests

Leaching tests are becoming increasingly important for assessing the risk of release of potential pollutants from a wide variety of solids (e.g. contaminated soils, mineral and demolition waste, construction products, etc.) [35-37]. Generally, the leaching tests can be performed as batch leaching tests and/or column leaching tests. Both test designs can be considered to have different possibilities and limitations depending on the purpose of the investigation. In general, batch tests are more simple compared to column tests which are more practically challenging. At the same time the column tests are better in mimicking the leaching process in the field. In the past, batch tests (or shaking tests) originally developed for sewage sludge, has also been used for organic compounds as sorption/desorption tests [35]. In Paper I, a series of batch leaching experiments were performed according to an experimental design to study the influence of the soil composition (proportion of peat, clay and sand) on the leachability of a wide range of HOCs. The tests were performed in 0.5 L Schott Duran bottles at a liquid to solid ratio (L/S) of 5 L/kg, which was reached by adding 0.5 L of an aqueous solution containing 0.001 M CaCl2 and 0.2

g/L NaN3 to each bottle (see Fig. 7). CaCl2 was added in the solution in order to adjust the ionic

strength and NaN3 to minimize risk for microbial degradation. The above mentioned

concentrations (0.001 M CaCl2 and 0.2 g/l NaN3) were used since these or similar one are

standardized in previous literature studies, for example Persson et al 2008 [2] and also on the international standard on leaching test (ISO/TS 21268-1:2007). The bottles were rotated at 120 rpm for 24 hours to obtain equilibrium between the contaminants in solution and contaminants in the soil.

In Paper II, a series of batch leaching experiments as pH static tests were performed with initial acid/base addition to investigate the influence of pH on the leachability of selected PCBs in naturally aged soil (Västervik, Sweden). The experiment setup was identically with the one used in Paper I, excepting the used of NaN3, the use of liquid to solid ratio (L/S) of 10 and the running

time of 48h. In addition, the pH of the leachates was adjusted to pre-selected set points in the range of pH 2 - 9 using HCl and NaOH solutions of 1, 0.1 and 0.01 M.

(24)

Fig. 7. A batch leaching test for organic compounds.

2.3.1. Factors affecting leachability

The leachability of contaminants in soil is dependent on the inherent physico-chemical properties of the compounds, the soil type, the presence of other contaminants, as well as the age of the contamination [35]. Basically, for hydrophobic organic compound there are three major factors that control the concentration in the aqueous phase in batch or column test systems (or natural soil systems): viz. 1) dissolution, 2) desorption, and 3) the presence of dissolved organic matter and colloids (organic and inorganic). If there is a free organic phase in the contaminated soil dissolution of the organic compounds from the free phase into the aqueous phase controls the concentration in the aqueous phase. If there is no free phase, the concentration in the aqueous phase is controlled by desorption and by the dissolved organic matter (DOM) and colloids. In both Paper I and II, the leachability of the target compounds was believe to be driven by desorption and by the dissolved organic matter (DOM) and colloids. Nevertheless, in Paper I, the leachability of the taget compounds was driven mostly by sorption and desorption of HOCs on

(25)

organic matter (peat) and clay, while the TOC was varying between 49.9 and 323 mg/L . In Paper II , the sorption and desorption of PCBs (due to the high content of clay and organic matter from the soil, see Tab. 2) can be still considered important. Nevertheles, in Paper II, the strong variation of TOC (from 62.4 to 465.0 mg/L) while the pH was varying from 2 to 9, can be considered the driving factor that controls the leachability of PCBs.

2.3.1.1. pH

pH is an important parameter, in relation to leaching of organic compounds from soil. This is partly due to dissociation of organic acids and to the fact that the DOC-concentration generally is strongly pH dependent and it is known that different fractions of organic acids will go into solution at different pH values, with fulvic acids being released at lower pH values and humic acids at pH 7 and higher. pH has been found to greatly affect the sorption of ionisable organic compounds, e.g. chlorophenols, since the pH affects not only the speciation of compounds but also the surface characteristics of natural solids (e.g. surface charge and potential). At neutral pH, the overall sorption has been found to be dominated by sorption of the dissociated fraction [5, 6]. In Paper I, the pH was estimated to vary between 7.70 in the case of M2 leachate (5 % peat) to 4.61 in the case of M4 leachate (60% peat), while concentration of PCP in the leachate was decreasing about 6 times. However, we concluded that this decrease cannot be entirely attributed to the hydrophobic interactions and in the case of slightly alkaline leachate (M2), the sorption of phenolate ion and the formation and sorption of the neutral metal-phenolate ion pair must be taken into account [18]. These changes of pH did not significant influence on sorption of the other non-ionisable organic compounds used in Paper I. In Paper II, the leachability of PCBs was mainly influenced by the increase of DOC, however the highly acid pH 2 and 3 of the leachates enable the preferential sorption of the least ortho-substituted PCB 66 and PCB 105.

2.3.1.2. Organic Matter (OM)

As mentioned earlier, the leachability of organic compounds in soil/sediments is suggested to be driven by the sorption/desorption and dissolution processes in relation with DOM [38]. With respect with the solubility of organic substances in the aqueous phases of the terrestrial

(26)

environment (soil and water) as well as of the aquatic environment, the natural organic material (NOM) can be divided into two categories: dissolved (DOM) and particulate (POM) organic material [39]. Dissolved organic matter (DOM) can act as a transport agent increasing the risk of spread

of a contaminant but on the other hand particulate organic matter (POM) can act as a trap decreasing potential mobilization [34, 37, 40-42]. In both Paper I and II, the DOM was defined as the experimentally defined fraction with diameters <0.7 μm since the DOM was separated from POM by a glass fiber filter with a pore size of 0.7 μm. However, the water phase is likely to contain colloids smaller than 0.7 μm, which may contribute to divergences from calculated log Kd values in Paper I. If colloid-facilitated transport is not taken into account the mobility and risk

of contaminants being transported from contaminated soils may be somewhat underestimated

[43].

2.4. Characterization of soil organic matter and matrices 2.4.1. X-ray Photoelectron Spectroscopy (XPS)

X-ray photoelectron spectroscopy (XPS) is a quantitative spectroscopic technique that measures the elemental composition, empirical formula, chemical state and electronic state of the elements that exist outermost 10 nm of a material’s surface. In XPS, the sample is bombarded with photons, electrons or ions to emit the photons, electrons or ions and then the kinetic energy of emitted electrons are measured. The binding energy of the electrons is characteristic of each element, and is also influenced by the chemical surroundings and oxidation state of the atom. Therefore XPS is widely used for determining the surface composition of different materials. In soil science XPS has been extensively used to characterized the gross C chemistry of soil organic matter [44].

In Paper I, X-ray photoelectron spectroscopy (XPS) was used to characterize the gross C chemistry of soil organic matter (peat) used to prepare the soils. Spectra were collected with an electron spectrometer (Kratos Axis Ultra DLD) using a monochromated Al Kα source operated

at 150 W (see Fig. 8). To compensate for surface charging, a low-energy electron gun was used. The binding energy (BE) scale was referenced to the C1s line of aliphatic carbon, set at 285.0 eV.

(27)

The spectra were processed using the manufacturer’s software (Kratos Analytical Ltd, Manchester, UK) Five binding chemical states of carbon were identified and quantified in the C1s

spectra: 285.0 eV (representing aliphatic C-C and C-H bonds), 285.7 (C-COOH, C-CON), 286.7 eV (C-O-C, C-OH, C-N), 288.2 eV (O-C-O, C=O) and 289.3 eV (COOH bonds). Regarding the other

Fig. 8. The Kratos Axis Ultra DLD XPS spectrometer.

elements, two components were identified for O1s (C=O at 531.7 eV and C-O at 533.1 eV) and

two for N1s ( non-protonated organic nitrogen at 400.4 eV and protonated organic nitrogen at

402.0 eV). The precision of atomic concentrations was approximately ±2 atomic %.

2.4.2. Fourier transform infrared spectroscopy (FTIR)

Since last decades, FTIR has been extensively used to in various application of soil science, e.g. sorption of organic compounds, characterization of humic substances and whole soils, structure

(28)

and reactivity of humic substances, studies of clay minerals, surface characterization of minerals, sorption/abiotic degradation of organic compounds, etc. [45]. Through vibration of their structure chemical bonds, FTIR can be used to distinguish the type of organic matter in soil such as carbohydrate, cellulose, and lipids [46]. One major advantage of FTIR over other surface techniques is that it does not require complicated and extensive sample preparation and the analysis takes only minutes. FTIR has been also used to characterize the organic fractions of DOM [47].

Fig 9. The FT-IR (BRUKAR IFS 66v/S) spectrometer used in Paper II.

Sample preparation for FTIR can be performed using different experimental techniques, freeze drying being one technique for studies of small amount of DOM in water samples [48].

In Paper II, FTIR was used to characterize the DOM obtained by freeze drying of the leachates. The leachates from soil were placed in a freezer for 24 hours at -200C. The frozen leachates were then freeze dried using a HETOSICC freeze drier machine. Pressure and temperature were maintained at 0.2-0.1 mPa and -400C to -500C respectively. Samples were completely dried in

(29)

powder form. After freeze drying, 4 to 10 mg of dry sample was mixed with potassium bromide (infrared spectroscopy grade, Fisher Scientific, Loughborough, UK) to a total weight of about 400 mg. They were ground and homogenized by agate pestle and mortar. Spectra were recorded over the 4000 – 400 cm-1 range, with 4 cm-1 resolution, under mild vacuum conditions (3 mbar) using a BRUKER IFS 66v/S instrument, equipped with a DTGS detector. 128 interferograms were co-added to obtain a good signal to noise ratio. Spectra were baseline corrected (64 point rubberband) and vector (area square sum) normalized using the software OPUS (version 5.5, Bruker Optics GmbH, Ettlingen, Germany) prior to multivariate analysis.

(30)

3. Results and Discussions

3.1. Variation of leachability in relation to soil composition and dissolved organic carbon

In Paper I, the log Kd varied between 2.0 for α-HCH in the soil with 90% sand, 5% clay and 5%

peat (M2), and 5.7 for PCB 155, the compound with the highest log Kow value (7.2), in the soil

with 30% sand, 60% clay and 10% peat (M6). For the compounds with lower hydrophobicity (described as log Kow), i.e. the HCHs, PCP, Phe and PCB 47, the highest Kd-values were

recorded for the soil with the highest proportion of organic matter (M4 with 20% sand, 20% clay and 60 % peat). For most of the model compounds (HCHs, PCP, HCB, PCB 47 and PCB 155), the lowest Kd-values were recorded for the soils with the lowest peat and clay contents (soils M1

and M2). In Paper II, the log Kd-values of all target PCBs, describing the their distribution

between leachates and soil (LOI of 37.5 %) decreased with increased pH values, the highest values being recorded at the initial pH ~2.0 and the lowest log Kd-values at initial pH ~9. This

correlation was stronger as the hydrophobicity of the PCBs increased, i.e. the correlation was weakest for PCB 28. The log Kd values for the PCBs were more different at low pH than at high

pH, meaning that the studied PCB congeners showed more similar “solubility” at high pHs, which might be due to their increased apparent solubility by highly values of TOC as a results of higher pH [49]. The increased levels of TOC in the leachates at higher pH thus seem to solubilised all PCB to a similar extent (see Tab.3). The highest compound-specific variation with pH of the log Kd-values with TOC was recorded for PCB 105 (the log Kd was decreasing from

6.8 to 3.25 comparing pH ~2 and ~9, respectively).

In Paper I, the lowest log Kd value (4.1) for PCB 153 was recorded for the M4 soil (20% sand,

20% clay and 60% peat) and the highest log Kd value (4.7) was recorded for soil containing 50 %

sand, 20 % clay, 30% peat (M3), while for the soil with 90% sand, 5% clay and 5% peat (M2), the log Kd of PCB 153 was 4.2. Further, in Paper I, the pH of the leachates was measured in a

separate DOC availability test using the same liquid to solid ratio (L/S) of 5 L/kg) was decreasing from pH 7.70 in the case of M2 leachate (5 % peat) to pH 4.61 in the case of M4 leachate (60% peat).

(31)

Tab.3. The average log Kd -values for target PCBs during the pH static test (Paper II). Compound pH 2 pH 3 pH 4 pH 5 pH 6 pH 7 pH 8 pH 9 Kow [50] PCB-28 5.71±0.09 5.47±0.01 5.72±0.03 5.57±0.13 5.63±0.15 4.25±0.06 3.83±0.01 3.34±0.11 5.67 PCB-52 5.89±0.09 5.71±0.01 6.03±0.05 5.92±0.02 4.53±0.14 4.20±0.11 3.76±0.01 3.22±0.12 5.84 PCB-66 6.00±0.10 5.79±0.03 5.98±0.05 5.88±0.02 4.54±0.13 4.19±0.09 3.80±0.01 3.25±0.10 6.20 PCB-101 5.97±0.07 5.92±0.01 6.19±0.06 6.08±0.02 4.55±0.13 4.19±0.10 3.75±0.01 3.24±0.11 6.38 PCB-105 6.08±0.08 5.92±0.02 6.18±0.06 6.13±0.06 4.59±0.12 4.17±0.09 3.73±0.00 3.25±0.11 6.65 PCB-118 5.93±0.08 5.78±0.01 6.01±0.06 5.92±0.03 4.39±0.12 4.16±0.09 3.71±0.00 3.21±0.10 6.74 PCB-138 5.68±0.08 5.78±0.02 6.01±0.07 5.92±0.02 4.39±0.12 4.16±0.10 3.71±0.02 3.21±0.08 6.83 PCB-153 5.71±0.09 5.84±0.04 6.00±0.09 5.94±0.06 4.45±0.12 4.16±0.11 3.72±0.00 3.22±0.07 6.72 PCB-156 6.56±0.13 6.62±0.01 6.76±0.07 6.70±0.01 5.10±0.13 4.73±0.08 4.43±0.01 3.93±0.07 7.18 PCB-180 5.68±0.10 5.91±0.08 6.03±0.09 5.98±0.01 4.52±0.12 4.15±0.10 3.71±0.02 3.20±0.07 7.21 PCB-187 5.87±0.17 6.05±0.02 6.25±0.09 6.14±0.03 4.49±0.12 4.14±0.10 3.68±0.02 3.20±0.07 7.17 Average initial pH 2.02±0.04 3.32±0.01 4.17±0.06 5.19±0.08 6.11±0.03 7.28±0.12 8.22±0.05 9.31±0.06 Average final pH 2.32±0.04 3.95±0.01 4.16±0.03 4.79±0.04 5.28±0.03 6.33±0.01 7.20±0.01 8.05±0.35 TOC (mg/L) 62.4±19.3 57.7 67.2±33.7 65.0±19.4 92.8±51.5 80.9±23.4 133.0±19.8 465.0±115.3

Therefore, in Paper I, we concluded that the pH did not influenced significantly the Kd-values

PCB 153. In contrast, in Paper II the Kd-values of PCB 153 are decreasing significantly with the

increase of pH, varying from 5.7 at pH of ~2 (initial value) to log Kd of 3.2 at the pH of ~9.

In Paper I, the relation between the soil composition (% sand, clay and peat, respectively), the physico-chemical properties of the model compounds, and the leachability of the same compounds (log Kd values), were investigated and explored with orthogonal projections to latent

structures (OPLS). The resulting OPLS scatter plot (Fig. 10) showed the degree to which the Kd

values for each compound were influenced by the soil constituents, but also for identification of compounds with similarly behavior in the system. In addition, compound specific response surfaces were calculated using screening models calculated with MODDE 9.0 software [51]. The

(32)

resulting contour plots show how the leachate concentrations of the compounds varied with the soil composition. The OPLS loading scatter plot (Fig. 10) indicate four groups of compounds with different interactions with the soil matrix.

For the first group (Group A in Fig. 10), consisting of HCHs, Phe and PCP, the Kd -values had a

relatively strong positive correlation with the peat content, while the correlation with the clay and the sand was weaker. For example, the response surface of α-HCH (Fig. 11) show that the concentration of α-HCH in the leachate is decreasing about 13 times, as the peat content increases from 0 to 60%.

The second group of compounds (Group B in Fig. 10) consists of TCDD and 1,3,6,8-TCDF. Their Kd-values were positively correlated with the peat content, but negatively

correlated with the clay content. Thus, 1,3,6,8-TCDD and 1,3,6,8-TCDF concentrations in the leachates decreased with increasing peat content in the soil, but increased with increasing clay content. This behavior for the 1,3,6,8-TCDD and 1,3,6,8-TCDF can also be seen in the contour plots for these compounds, shown in Figs. 12A and 12B, respectively. The contour plots are similar, indicating that these two compounds have similar leaching behavior, and that they were affected similarly by the studied soil components.

For the 3rd group of compounds (Group C in Fig. 10), consisting of BaA, HCB and PCB 47, the K

d values

were positively correlated with both the peat and the clay content, meaning that the leachate concentrations of these compounds were decreasing as both the peat and the clay content increased. However, compared to Group A, the correlation with the peat content was weaker for Group C, and for BaA even the correlation with clay was weak. For example, the contour plot presented in Fig. 13 combined with Fig. 10 shows that the concentration of HCB in leachates is increasing with the decrease of both peat and clay content of the soil as we are moving from the bottom of plot (high content of clay and peat) to the top (low content of clay and peat).

(33)

-0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 pq[ 2] pq[1] a-HCH b-HCH g-HCH d-HCH PCP HCB BaA Phe PCB-47 PCB-153 PCB-155 TCDF TCDD Sand % Clay % Peat % D B A C

Fig. 10. OPLS loading scatter plots of Kd values for the model compounds.

Fig. 11. Contour plot showing the variation of concentrations in leachates of α-HCH vs. proportions of

(34)

A

B

Fig. 12. Contour plots showing the variation of concentrations in leachates of 1,3,6,8-TCDD (A) and

1,3,6,8-TCDF (B) vs. proportions of clay and peat in the artificial soils. The R2 values of the MODDE

screening models were 0.68 for 1,3,6,8-TCDD and 0.68 for 1,3,6,8-TCDF respectively.

The 4th group of compounds (Group D in Fig. 10), consisting of PCB 153 and PCB 155, showed Kd values that, according to the OPLS-model, were positively correlated with the clay content

while being negatively correlated with the peat content. This means that these compounds would be expected to show higher leachability as the clay content decreased, but also as the peat content increased , for PCB 153, see Fig.14.

(35)

 

Fig. 13. Contour plots showing the variation of concentrations in leachates of HCB vs. proportions of clay

and peat in the artificial soils. The R2 values of the MODDE screening model was 0.69.

Fig. 14. Contour plot showing the variation of concentrations in leachates PCB 153 vs. proportions of

(36)

3.2. Qualitatively characterization of soil organic matter and dissolved organic carbon

In Paper I, X-ray photoelectron spectroscopy (XPS) was used to characterize the gross carbon chemistry of soil organic matter (peat) used to prepare the soils. Spectra were collected with an electron spectrometer (Kratos Axis Ultra DLD) using a monochromated Al Kα source operated

at 150 W. Five binding chemical states of carbon were identified and quantified in the C1s

spectra: 285.0 eV (representing aliphatic C-C and C-H bonds), 285.7 (C-COOH, C-CON), 286.7 eV (C-O-C, C-OH, C-N), 288.2 eV (O-C-O, C=O) and 289.3 eV (COOH bonds) (see Fig.15). Regarding the other elements, two components were identified for O1s (C=O at 531.7 eV and

C-O at 533.1 eV) and two for N1s (non-protonated organic nitrogen at 400.4 eV and protonated

organic nitrogen at 402.0 eV). By the XPS analysis, the following moieties were found: C, C-H, COOC-H, C-CON, C-O-C, C-OC-H, C-N, O-C-O, C=O (see fig. 15). Among them, the aliphatic moieties identified at 285.0 eV (C-C and C-H bonds) have the highest atomic concentrations (AC), 41.91 %, followed by the components identified at the binding energy of 286.7 eV (C-OH, C-N and C-O-C bonds) with 20.5 %. The high presence of aliphatic moieties in the peat combined with the high influence of the peat content on the Kd-values of HCHs, Phe and PCP

suggests that alkyl-C moieties, in addition to aromatic moieties, can act as sorption domains for HOCs in soil organic matter. The results indicate that, the alkyl-C moieties can sorb appreciable amounts of HOCs, in some cases with even higher affinity than aromatic-rich sorbents as report by Chefetz et al. in a review article [52]. The above mentioned review article concluded that neither aromaticity nor aliphaticity of SOM alone can be used to predict the sorption affinity of sorbents having wide and diverse properties and therefore the aliphatic structures must be considered in the evaluation of HOC-sorption processes in the environment. More specifically they [52]. concluded that phenanthrene has a strong affinity for aliphatic SOM domains.

In Paper II, the composition of dissolved organic matter (DOM) at all pH values (except the DOM obtained at the pH ~2) was explored by FTIR spectrometry. Further, in order to explore the correlation between dissolved organic matter and leachability of PCBs, the relationship between the log-Kd of the target PCBs and digitalized FTIR spectra were explored by orthogonal projection to latent structures (OPLS).

(37)

Fig. 15. The high resolution XPS spectra of carbon states recorded for the peat at 285.0 eV (representing

aliphatic C-C and C-H bonds), 285.7 (C-COOH, C-CON), 286.7 eV (C-O-C, C-OH, C-N), 288.2 eV (O-C-O, C=O) and 289.3 eV (COOH bonds).

An one-dimensional model with N=17 observations and K=1671 variables (X=1660, Y=11) was obtained. In order to demonstrate the influence of pH on the FTIR spectra of DOM, the OPLS loadings were used to describe all bands (peaks) in relation with the FTIR spectra of two extreme samples (pH ~3 vs. pH ~9, Fig. 3), focusing on similarities and differences in spectra.

(38)

Fig. 16. Co-relation of OPLS Loadings plot for PC1 (upper) and FTIR spectra (lower) of DOM at pH 3

(blue) and pH 9 (green).

Peak intensity and sign (positive or negative) in loadings help to identify regions of importance in the FTIR spectra related to the differences in Kd. Peaks at 3558 and 3414 cm-1 were the most

intense on the negative side of predictive component. These peaks may attribute to O-H stretching and/or N-H stretching. Therefore we can conclude that that O-H and/or N-H functional groups are showing higher bands at pH ~3 than at pH ~9. On the other hand, peaks at 1155, 1591 and 1670 cm-1 showed highest bands in the positive side of the loadings. These peaks can originate from C-O stretching and/or O-H deformation, aromatic C=C stretching and C=O stretching vibrations, respectively. Thus, we can conclude that these peaks showed higher bands at pH ~9 than at pH ~3. The peak recorded at 1036 cm-1 showed the highest band in the positive

side of predictive component resulting from the spectra recorded at pH ~9 while this peak was very weak in spectra recorded at pH ~3. This peak can be related to the C-O stretching of alcohol, ethers [53] and/or polysaccharides [54]. C-H stretching and N-H group may be responsible for peaks at 1531 cm-1 and 2935 cm-1 in the loadings plot. These peaks resulted from

(39)

the spectra recorded at pH ~9 because they were not present in spectra recorded at pH ~3. Another difference was also observed in the region of 600-700 cm-1 where two sharp peaks are produced from two different samples. Since these bands originate from composite vibrations, their assignment was less straightforward. Nevertheless, the differences detected in the above mentioned functional groups demonstrated that the composition and concentration of DOM in the leachates were significantly influenced by the change of pH of the leachates as moving from an acidic to an akaline domain.

3.3. Variation of the distribution coefficients with hydrophobicity of the compounds

In Paper I, the log Kd-values of all compounds calculated for OECD soil as well for the soils

with the highest amount of peat (M4) and clay (M6), respectively, were plotted against their log Kow (See Fig. 17). The data from the log Kd vs. log Kow plot for OECD soil (Fig. 17A) were fitted

linearly and also quadratic, and the slope of the linear curve was calculated to 1.02 ± 0.18 (Eqn. 1 from Fig.17A), close to the value of 1, as expected for sorption processes that are driven by hydrophobic interactions [55]. However, for the more hydrophobic compounds (log Kow>6) [56]

the measured Kd-values deviated from the linear relationship, which also the regression

coefficient (R2), calculated to 0.71, indicate. Excluding the Kd values for the four most

hydrophobic compounds (PCB 153 and PCB 155, 1,3,6,8-TCDD and 1,3,6,8-TCDF) a better linear relationship was calculated (R2 = 0.85), while the slope of the linear curve was 1.34 ± 0.19 (Eqn.2 from Fig.17A). The log Kd vs. log Kow plots for the more extreme soils in terms of peat

and clay content M4 and M6 soils shown in Figs. 17B and 17C, indicate an even more complex relationship. Excluding the Kd values for the four most hydrophobic compounds (PCB 153 and

PCB 155, 1,3,6,8-TCDD and 1,3,6,8-TCDF), the linear fitting of the all Kd data for both M4 and

M6 soils was resulting much more low R2 values (data not shown) than the one calculated with all Kd data for OECD soil. The slopes of the curves for log Kd vs. log Kow plots for the M4 and

M6 soils were 0.65 ± 0.18 for M4 and 1.47 ± 0.17 for M6 respectively, clearly deviating from the suggested value of 1 [55]. Thus log Kd vs. log Kow plots from Fig. 17 indicated that the

apparent Kd values show a deviation from the “true Kd values” for highly hydrophobic

(40)

3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 1 2 3 4 5 6 7 8 9 10 11 12 13 B y = 0.6578x + 1.4665 R2 =0.58 log K d log Kow 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 1 2 3 4 5 6 7 8 9 10 11 12 13 C y = 1.4793x - 2.97163 R2 =0.90 log K d log Kow 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 1 2 3 4 5 6 7 8 9 10 11 12 13 A y = 1.0266x - 1.0291 (1) R2 =0.71 y = 1.3478x - 2.4133 (2) R2 =0.85 log K d log Kow

Fig. 17. Log Kd values for the target compounds vs. their log Kow values,calculated for OECD standard

soil (A), M4 soil (20% sand, 20% clay and 60 % peat) (B) and M6 soil (30 % sand, 60 % clay, 10% peat) (C). The OECD data were fitted linearly (black curve for all data and green curve excluding compounds 10 to 13) and quadratically (red curve for all data) (A). Eqns. (1) and (2) describe the linear fits of all data and data excluding compounds 10 to 13, respectively (A). The target compounds were -HCH (1), -HCH (2), --HCH (3), --HCH (4), PCP (5), HCB (6), BaA (7), Phe (8), PCB 47 (9), PCB 153 (10), PCB 155 (11), 1,3,6,8-TCDF (12) and 1,3,6,8-TCDD (13).

In Paper II, the the log Kd-values of all target PCBs recorded at the pH ~3, pH~6 and pH~9 were

plotted against their log Kow (see Fig. 18). The data from the log Kd vs. log Kow plot for Kd values

recorded at pH ~3 were fitted linearly and the slope of the linear curve was calculated to 0.53± 0.12 and the regression coefficient was calculated (R2=0.62), indicating a the linear relationship,

(41)

as expected for sorption processes that are driven by hydrophobic interactions [55]. The data from the log Kd vs. log Kow plot for Kd values recorded both at pH~6 and pH ~9 do not followed

the same linearity since the PCBs have more similar log Kd values at higher pH, which might be

due to their increased apparent solubility by highly values of TOC [49].

5.50 5.75 6.00 6.25 6.50 6.75 7.00 7.25 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 y = 0.1356x + 2.4043 R2=-0.04 y = -0.2316x + 6.1583 R2=0.01 y = 0.5332x + 2.4701 R2=0.62

lo

g

K

d

of

t

raget

PCBs

log Kow

Fig. 18. Log Kd values for the target PCBs vs. their log Kow values,calculated at pH ~3 (■), pH ~6 (▲)

and pH~9 (●).

3.4. Correlation between leachability of ortho-PCBs and DOC

In order to investigate the influence of molecular structure of selected ortho-PCBs on their leachability the fractions of concentrations of mono-ortho to di-ortho PCBs (PCB 66 (2,3’,4,4’) to PCB 52 (2,2’,5,5’), PCB 105 (2,3,3’,4,4’) to PCB 101 (2,2’,4,5,5’), and PCB 156 (2,3,3’,4,4’,5) to PCB 153 (2,2’,4,4’,5,5’) and the fractions of concentrations of di-ortho to

(42)

tri-ortho PCBs (PCB 180 (2,2’,3,4,4’,5,5’) to PCB 187 (2,2’,3,4’,5,5’,6) were plotted vs. pH and TOC (see Fig. 19).

1 2 3 4 5 6 7 8 9 10 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 pH Fraction of PCBs Concentration 0 100 200 300 400 500 600 TOC (mg/ L)

Fig. 19. Change in fractions of concentration of PCB 66 to PCB 52 (●), PCB 105 to PCB 101 (▲), PCB

156 to PCB 153 (▼) and PCB 180 to PCB 187 (■) respectively and total organic carbon (□) vs. pH.

The fractions of concentrations of tetrachlorobiphenyl congeners PCB 66 (mono-ortho) to PCB 52 (di-ortho) increased from 0.451 ± 0.012 (pH ~2) to 0.541±0.001 pH (pH ~4) and afterwards decreased to 0.497±0.016 at pH (pH ~9) indicating an influence of the ortho-chlorine atoms through pH effects on the leachability of PCB 52. Also the fractions of concentrations of pentachlorobiphenyl congeners PCB 105 (mono-ortho) to PCB 101 (di-ortho) increased from 0.328 ± 0.017 (pH ~2) to 0.391±0.002 pH (pH ~4) and afterwards remaining relatively constant to 0.384± 0.006 at pH (pH ~9). In contrast, the fraction of PCB 156 (mono-ortho) to PCB 153 (di-ortho) varied from 0.082 ± 0.006 (pH ~2) to 0.098 ± 0.016 pH (pH ~4) and afterwards to

(43)

0.109± 0.004 (pH ~9) without any clearly trend. The fraction of concentrations of bulky heptachlorobiphenyl congeners PCB 180 (di-ortho) to PCB 187 (tri-ortho) varied from 0.759 ± 0.002 (pH ~2) to 0.773 ± 0.02 (pH ~4) without a clearly trend and afterwards decreased to 0.690± 0.005 (pH ~9) clearly showing a more complex influence of the ortho-chlorine atoms on the leachability of PCB 187.

(44)

4. Concluding remarks and future aspects

4.1. Conclusions

The present thesis has demonstrated how the complex interaction between both the organic matter and clay components influences the leachability of highly hydrophobic compounds in a compound-specific manner. It was shown that the leachability of a broad range of model compounds was driven by sorption/desorption mechanisms involving both the fractions of organic matter and clay. Generally, for all model compounds studies, the Kd-values showed a

variability of 2-3 orders of magnitude depending on the matrix composition. More specifically the Kd-values of moderately hydrophobic compounds, such as HCHs, PCP and Phe, were

correlated mainly with the organic matter content of soil, and their leachability decreases as the peat content increases, while the correlation with the clay content was much weaker. For slightly more hydrophobic compounds, such as BaA, HCB and PCB 47, there were correlations with both OM and clay contents implying that their leachability decreases as the proportions of either of these soil constituents increases.

Interestingly, a deviation behavior of two classes of highly hydrophobic classes of compounds was observed. For the PCDDs/Fs, i.e. 1,3,6,8-TCDD and 1,3,6,8-TCDF, the leachability was negatively correlated with the peat content but positively correlated with the clay content, while for the highly chlorinated PCBs, i.e. PCB 153 and PCB 155, the correlations were the reversed The relationship log Kd vs. log Kow investigated showed that the apparent log Kd values for

highly hydrophobic compounds may deviate from the linear relationship seen for the less hydrophobic compounds due to colloids, present in the leachate.

The understanding of co-transport of contaminants as colloid inorganic or organic components via release will thus be of crucial importance and for highly hydrophobic compounds (Kow > 6),

more important than solubility in the surrounding media. Therefore, it was investigated how these processes were influenced by pH. At lower pH values, that leachability of PCBs was driven by the hydrophobic interactions. However, the log Kd values of all PCBs varied more at lower

pH than at higher pH which means that they have more similar “solubility” at high pH values, suggesting the importance of co-transport with increased release of DOC as a results of higher pH. The the tendency to leach was higher among the least chlorinated and lipophilic PCB

(45)

congeners (PCB 28 and PCB 52). The variation of fractions of concentrations of mono-ortho to di-ortho PCBs (PCB 66 to PCB 52 and PCB 105 to PCB 101) was also indicating a influence of the ortho-chlorine atoms through pH effects on their leachability. We suggested that the decreased fraction of concentrations recorded at the pH~2 of the leachate might be cause by comparative difficulty of more ortho-subsititued PCBs to achieve a coplanar ring conformation due to steric effects of the ortho-chlorines. The FTIR characterization of the DOC-fractions at different pH-values demonstrated that the composition and concentration of dissolved organic matter (DOM) at different pH clearly influenced the leachability of different individual PCBs. Some compound-specific correlations between Kd-value and DOC characteristics could be

identified. The results of the present thesis contribute to the understanding and assessment of complex processes connected to the mobility of organic contaminants in soil:  

4.2. Future aspects

In the area of risk assessment of contaminants in soil much focus has been put on the compounds itself and less on the system and processes determining the transport and fate in the soil system. The organic matter has generally been considered to be important, mainly as the bulk to which hydrophobic compounds associate. Beside the organic matter content of the soil, the clay content may also influence the leachability of HOCs, especially for highly hydrophobic compounds (Kow

> 6), we are suggesting that specific surface to the clay to be investigated in further studies. Moreover, a corresponding mechanism discussion on different groups of compounds behavior on clay or peat is lacking and need elaboration. The influence of the ortho-chlorine bonds through pH effects on leachability PCBs should be explored in further studies using a variety of model congeners (i.e. PCB 169 vs. PCB 155). The pH effects on leachability of HOCs should be further explored for new classes of compounds (for example polybrominated biphenyls (PBBs vs. PCBs due to the change of halogen atoms). For the better understanding of leachability of HOCs and the possible interactions with different pools of organic matter, the dissolved organic carbon (DOC) should be better explored by a variety of analytical techniques (FTIR, XPS, Dynamic Light Scattering (for colloids), LC-Orbitrap MS (for characterization of low molecular humic acids). Furthermore, since it is not yet understood whether sorption of contaminants to soil components is linked to significant isotope fractionation, these isotope effect should be

(46)

investigated in further studies. Since the increase in the isotope shift is proportional to the sorption tendency (Koc value) and in the case of carbon and hydrogen isotopes, the heavy

isotopomers move faster than light ones, we expect that the isotope composition of the contaminates in the soil to be clearly lighter comparing with the beginning of the leaching test and this isotope fractionation to decrease with time. As the isotope fractionation tends to increase as a function of the hydrophobicity of the analyte [57] we expect a higher fractionation factors for very hydrophobic contaminates (for example octachlorodibenzo-p-dioxin).

(47)

Acknowledgements

Although a thesis is basically the work of one single person, I would like to express my gratitude to many colleagues and friends without which this thesis would not had been possible. First, I would like to express my deepest gratitude to my supervisors, Mats and Staffan. Thanks for your inspiring devotion to science and for giving me the opportunity to graduate. Thank you Per for your personal support with the analytics and ideas at Environmental Chemistry lab. In science world, the big goals are reached only with small experimental steps and these small steps would not had been possible without you, Maria, Rolf and Sture.

Warm thanks to all my colleagues in the Environmental Chemistry 7th floor corridor, from the past and present (Sarah, Kristina A, Kristina S, Kristina F, Tomas, Eva, Malin, Matyas, Marcus, 

Aleksandra, Lan, Mehdi, Mandana, Edo Mar and Jin).

Thank you Majid, for your held in two lab experiments that will end up in two peer review publications.

Special thanks to Andrey Shchukarev for his support with XPS analysis and to András Gorzsás for his support with FTIR analysis. Thank you Rui, for your help and advices in multivariate data analysis.

SPECIAL thanks (notice the capital letters! ) to Chau. You’re a wonderful friend, always willing to help me when I needed (which was quite often, I have to say).

I would like to thank to Kempe Memorial Scholarship Foundation for their financial support and Wallenberg Foundation for awarding me two travel grants.

Least but not last, without my wonderful and supportive parents, I would not have reached this goal. Thank you for giving me the strength to go on when times were tough and supported me to get an education that you didn’t had the chance to get it in spite of your skills.

(48)

References

1. Jonsson S (2009) The influence of soil and contaminant properties on the efficiency of physical and chemical soil remediation methods. Doctoral thesis, University of Umeå, Umeå, Sweden.

2. Wang, X.; Lu, J.; Xu, M.; Xing, B., Sorption of Pyrene by Regular and Nanoscaled Metal Oxide Particles: Influence of Adsorbed Organic Matter. Environ. Sci. Technol. 2008, 42, (19), 7267-7272.

3. Kile, D. E.; Chiou, C. T.; Zhou, H.; Li, H.; Xu, O., Partition of Nonpolar Organic Pollutants from Water to Soil and Sediment Organic Matters. Environ. Sci. Technol. 1995, 29, (5), 1401-1406.

4. Schwarzenbach, R. P.; Westall, J., Transport of nonpolar organic compounds from surface water to groundwater. Laboratory sorption studies. Environ. Sci. Technol. 1981, 15, (11), 1360-1367.

5. Jacobsen, B. N.; Arvin, E.; Reinders, M., Factors affecting sorption of pentachlorophenol to suspended microbial biomass. Water Res. 1996, 30, (1), 13-20.

6. Shimizu, Y.; Yamazaki, S.; Terashima, Y., Sorption of anionic pentachlorophenol (PCP) in aquatic environments: The effect of pH. Water Sci. Technol. 1992, 25, (11), 41-48.

7. Bronner, G.; Goss, K.-U., Sorption of Organic Chemicals to Soil Organic Matter: Influence of Soil Variability and pH Dependence. Environ. Sci. Technol. 2010, 45, (4), 1307-1312.

8. Willett, K. L.; Ulrich, E. M.; Hites, R. A., Differential Toxicity and Environmental Fates of Hexachlorocyclohexane Isomers. Environ. Sci. Technol. 1998, 32, (15), 2197-2207.

9. Walker, K.; Vallero, D. A.; Lewis, R. G., Factors influencing the distribution of lindane and other hexachlorocyclohexanes in the environment. Environ. Sci. Technol. 1999, 33, (24), 4373-4378.

10. Phillips, T.; Seech, A.; Lee, H.; Trevors, J., Biodegradation of hexachlorocyclohexane (HCH) by microorganisms. Biodegradation 2005, 16, (4), 363-392.

(49)

11. Hollifield, H. C., Rapid Nephelometric Estimate of Water Solubility of Highly Insoluble Organic-Chemicals of Environmental Interest. Bull. Environ. Contam. Toxicol. 1979, 23, (4-5), 579-586.

12. Melancon, S. M.; Pollard, J. E.; Hern, S. C., Evaluation of Sesoil, Przm and Pestan in a Laboratory Column Leaching Experiment. Environ. Toxicol. Chem. 1986, 5, (10), 865-878. 13. Reinhart, D. R.; Pohland, F. G., The Assimilation of Organic Hazardous Wastes by Municipal Solid-Waste Landfills. J. Ind. Microbiol. 1991, 8, (3), 193-200.

14. Nordmeyer, H.; Pestemer, W.; Rahman, A., Sorption and Transport Behavior of Some Pesticides in Groundwater Sediments. Stygologia 1992, 7, (1), 3-11.

15. Wiberg K. 2002. Enantiospecific analysis and environmental behavior of chiral persistent organic pollutants (POPs). PhD thesis. Environmental Chemistry, Department of Chemistry, Umeå University, Umeå, Sweden. ISBN 91‐7305‐162‐4. pp 1‐70. In.

16. Han, J.; Deming, R. L.; Tao, F. M., Theoretical study of molecular structures and properties of the complete series of chlorophenols. J. Phys. Chem. A 2004, 108, (38), 7736-7743. 17. Wahlström, M.; Laine-Ylijoki, J.; Pihlajaniemi, M.; Ojala, M., Leaching of PCBs and chlorophenols from contaminated soil and waste-influence of leaching test characteristics. In Waste Manage. Ser., G.R. Woolley, J. J. J. M. G.; Wainwright, P. J., Eds. Elsevier: 2000; Vol. Volume 1, pp 462-474.

18. Lee, L. S.; Rao, P. S. C.; Nkedi-Kizza, P.; Delfino, J. J., Influence of solvent and sorbent characteristics on distribution of pentachlorophenol in octanol-water and soil-water systems. Environ. Sci. Technol. 1990, 24, (5), 654-661.

19. United Nations Environment Programme (UNEP). (2001). Stockholm convention on Persistent Organic Pollutants (POPs).

20. Bernes,C. (1998). Persistent Organic Pollutants. Swedish Environmental Protection Agency.

21. Blumer, M., Polycyclic Aromatic-Compounds in Nature. Sci. Am. 1976, 234, (3), 35-45. 22. Khadhar, S.; Higashi, T.; Hamdi, H.; Matsuyama, S.; Charef, A., Distribution of 16 EPA-priority polycyclic aromatic hydrocarbons (PAHs) in sludges collected from nine Tunisian wastewater treatment plants. J. Hazard. Mater. 2010, 183, (1–3), 98-102.

(50)

23. Lundstedt S (2003) Analysis of PAHs and Their Transformation Products in Contaminated Soil and Remedial Processes. Doctoral thesis, University of Umeå, Umeå, Sweden 24. Ballschmiter K., Z., M., 1980. , Analysis of polychlorinated biphenyls (PCB) by glass capillary gas chromatography. Fresenius J. Anal. Chem. 1980.

25. Baars, A. J.; Bakker, M. I.; Baumann, R. A.; Boon, P. E.; Freijer, J. I.; Hoogenboom, L. A. P.; Hoogerbrugge, R.; van Klaveren, J. D.; Liem, A. K. D.; Traag, W. A.; de Vries, J., Dioxins, dioxin-like PCBs and non-dioxin-like PCBs in foodstuffs: occurrence and dietary intake in The Netherlands. Toxicol. Lett. 2004, 151, (1), 51-61.

26. Van den Berg, M.; Birnbaum, L. S.; Denison, M.; De Vito, M.; Farland, W.; Feeley, M.; Fiedler, H.; Hakansson, H.; Hanberg, A.; Haws, L.; Rose, M.; Safe, S.; Schrenk, D.; Tohyama, C.; Tritscher, A.; Tuomisto, J.; Tysklind, M.; Walker, N.; Peterson, R. E., The 2005 World Health Organization reevaluation of human and mammalian toxic equivalency factors for dioxins and dioxin-like compounds. Toxicol. Sci. 2006, 93, (2), 223-241.

27. Hawker, D. W.; Connell, D. W., Octanol Water Partition-Coefficients of Polychlorinated Biphenyl Congeners. Environ. Sci. Technol. 1988, 22, (4), 382-387.

28. Huang, Q. D.; Hong, C. S., Aqueous solubilities of non-ortho and mono-ortho PCBs at four temperatures. Water Res. 2002, 36, (14), 3543-3552.

29. Quass, U.; Fermann, M.; Broker, G., The European dioxin air emission inventory project - Final results. Chemosphere 2004, 54, (9), 1319-1327.

30. Fiedler, H., National PCDD/PCDF release inventories under the Stockholm convention on persistent organic pollutants. Chemosphere 2007, 67, (9), S96-S108.

31. Weber, R.; Gaus, C.; Tysklind, M.; Johnston, P.; Forter, M.; Hollert, H.; Heinisch, E.; Holoubek, I.; Lloyd-Smith, M.; Masunaga, S.; Moccarelli, P.; Santillo, D.; Seike, N.; Symons, R.; Torres, J. P. M.; Verta, M.; Varbelow, G.; Vijgen, J.; Watson, A.; Costner, P.; Woelz, J.; Wycisk, P.; Zennegg, M., Dioxin- and POP-contaminated sites-contemporary and future relevance and challenges. Environ. Sci. Pollut. Res. 2008, 15, (5), 363-393.

32. OECD, (1984). Organization for Economic Co-operation and Development. Test 207: earthworm, acute toxicity tests. In: Organization for Economic Co-operation and Development (ed.), OECD Guidelines for Testing of Chemicals.

References

Related documents

When soaking extraction of outsole material using synthetic brackish water was performed, as an attempt to model weathering and leaching in the environment, eight different

It is important to maintain high catalyst activity throughout its life cycle for optimal hydrotreater performance. During normal operation, loss of catalyst activity is inevitable.

For all created soil indices (Distance30, Depth4 and CArea) significant differences could be seen between points in the sedimentary (Sed) part and moraine (Till) part of

In Batch 5, a final attempt of extracting with water was done as well as extracting with a 50:50 (v/v) water/methanol mixture. Extraction was performed after soaking the soil

In human experimental exposure studies, symptoms of irritation have appeared at MVOC concentrations several orders of magnitude higher than those measured under field conditions

I de jämförelser som görs av de båda fallen kommer det fram att mordet på Palme inte var en engångsföreteelse utan i Sverige får befolkningen och framför

En utbildning av urbana matproducenter skulle därför kunna vara ett sätt att bryta denna onda cirkel och öka intresset hos människor att börja med kommersiellt urbant jordbruk..

However, employing the inherent structure of some important problems, such as H 1 and gain-scheduling synthesis, convexity can be recovered and the existence of a controller K can